Government accused of dragging its heels on deepfake law over Grok AI

Campaigners have accused the federal government of dragging its heels on implementing a regulation which might make it unlawful to create non-consensual sexualised deepfakes.

It comes amid a backlash in opposition to pictures created utilizing Elon Musk’s AI Grok to digitally take away clothes – with one girl telling the BBC greater than 100 sexualised pictures have been created of her.

It’s at present illegal to share deepfakes of adults in the UK, however new laws that might make it a legal offence to create or request them remains to be not in pressure regardless of passing in June 2025.

However it’s unclear whether or not the entire unclothing pictures created by Grok would fall foul of this regulation. The BBC has contacted the federal government for remark.

In a statement, X mentioned: “Anybody utilizing or prompting Grok to make unlawful content material will endure the identical penalties as in the event that they add unlawful content material.”

“We take motion in opposition to unlawful content material on X, together with Youngster Sexual Abuse Materials (CSAM), by eradicating it, completely suspending accounts, and dealing with native governments and regulation enforcement as essential,” the assertion mentioned.

Grok might be accessed by way of its web site and app, or by tagging “@grok” in posts on the social media platform X.

However Andrea Simon from Finish Violence Towards Ladies mentioned by not placing the regulation into impact, the federal government had “put ladies and women in hurt’s means”.

“Non-consensual sexually express deepfakes are a transparent violation of girls’s rights and have a long-lasting, traumatic influence on victims,” she mentioned.

“For girls utilizing platforms like X, the specter of this abuse may imply they really feel the necessity to self-censor and alter their behaviour, limiting their freedom of expression and participation on-line.”

“This isn’t solely a legal justice challenge however a difficulty of regulating a tech ecosystem that facilitates and income from violence in opposition to ladies and women.”

On Tuesday, Know-how Secretary Liz Kendall demanded X “cope with this urgently”, calling the scenario “completely appalling”.

The regulator Ofcom mentioned it had made “pressing contact” with X and xAI, which constructed Grok, on Monday and informed the BBC it was investigating considerations.

Each Kendall and Downing Avenue have backed the regulator taking motion, with the Prime Minister’s spokesperson including on Wednesday “all choices remained on the desk”.

The Ministry of Justice informed the BBC: “It’s already an offence to share intimate pictures on social media, together with deepfakes, with out consent.

“We refuse to tolerate this degrading and dangerous behaviour, which is why now we have additionally launched laws to ban their creation with out consent.”

The present regulation says it’s unlawful to generate pornographic deepfakes when utilized in revenge porn or depicting kids.

A provision within the Knowledge (Use and Entry) Act 2025 criminalised the creation or commissioning of “purported intimate pictures” – or deepfakes – Lorna Woods, Professor of Web regulation at Essex College, informed the BBC.

However a year on from first announcing a crackdown on the offence, the federal government has nonetheless not introduced a key authorized provision that might goal these asking grok to make sexualised deepfakes of girls and women, specialists and campaigners have informed the BBC.

Professor Woods mentioned this offence “would appear to be a superb match for a number of the pictures which have been created utilizing Grok” however it had not but been introduced into pressure.

In the meantime the EVAW’s Simon questioned why the secondary laws wanted to make creating sexual deepfakes against the law had nonetheless not occurred, given they have been “a transparent violation of girls’s rights”.

“This regulation has nonetheless not come into pressure,” she mentioned, “nor has a date been set for when it will happen”.

Conservative peer Baroness Owen, who campaigned for the legal change in the House of Lords, informed the BBC the federal government had “repeatedly dragged its heels” to convey the principles into impact.

“We can not afford any extra delays,” she mentioned.

“Survivors of this abuse deserve higher.

“Nobody ought to must dwell in concern of their consent being violated on this appalling means.”

And cross-bench peer Baroness Beeban Kidron informed the BBC: “Know-how strikes quick, and this laws is meant to plug an present hole, so there isn’t a excuse for delay.”

The BBC has spoken to several women who’ve had footage they posted of themselves on X became deepfakes by the Grok AI.

Grok is a free software which customers sometimes tag beneath different individuals’s feedback with prompts to offer a response or context.

Extra lately, individuals have been in a position to make use of the chatbot to change footage they or different individuals have uploaded by tagging it and together with a textual content immediate – even when that features undressing them or placing them in sexualised poses.

Evie informed the BBC whereas she observed individuals began to answer to footage of herself asking Grok to place her in a bikini a couple of months in the past, the most recent replace had made it “simpler” for customers to do, and the photographs seemed “extra practical”.

She says she now has had at the least 100 sexualised pictures of herself created by way of the AI assistant, so many in order that she has given up reporting all of them as a result of “psychological pressure” concerned in viewing them.

In addition to having to have a look at the pictures herself, the concept of family members seeing her in sexual or undressed poses has additionally made staying on X very tough for her.

“My household observe me on there, my mates, my co-workers,” she mentioned.

“Figuring out that each one the individuals I care about in my life can see me like that… it is disgusting.”

Dr Daisy Dixon is one other feminine X consumer who informed the BBC she has seen a rise in individuals utilizing Grok to undress her, significantly utilizing her profile image.

She mentioned the photographs left her feeling “humiliated” – and mentioned the actual fact Grok routinely commented the altered pictures at her left her feeling as if she had been assaulted.

“To have that energy transfer of posting it again to you – it is like saying ‘I’ve management over you and I’ll preserve reminding you I’ve management over you’,” she mentioned.

“We do not wish to dilute the idea, however it looks like a sort of assault on the physique.”

And X customers reminiscent of Evie say motion must occur now.

“There’s so many locations on-line that you are able to do this, however the truth that it was taking place on Twitter with the in-built AI bot – that is loopy that is allowed,” she mentioned.

“Why is that this allowed and why is nothing being finished about it?”